HDX of VSVI

The mechanism for gas phase HDX of negatively charged nucleic acids has been studied previously.³¹ A few mechanisms have been proposed, most of which involve the backbone phosphates interacting with the deuterated solvent.³¹ Following HDX, the mass spectra for each charge state was recorded and the relative rates of deuterium uptake were observed, as illustrated in Figure 4.6. Only the −6 charge state for each chemical modifier is depicted in Figure 4.6, while the HDX spectra for rest of the investigated charge states can be found in Appendix B.

Figure 4.6 also depicts the difference spectrum for each modifier which is the resultant spectrum when the non-HDX data is subtracted from HDX data and indicates where signal is lost and gained during HDX relative to deuterium uptake. It appears that solvent modifiers had a major effect on deuterium exchange with acetone enhancing uptake and IPA impeding uptake. The overall order for gaseous environments and deuterium exchange from most exchange to least was: acetone ≈ acetonitrile > nitrogen > isopropyl alcohol. Based on the chemical properties of the solvent modifiers, the polarity did not seem to play a role in exchange but the protic solvent modifiers displayed less exchange while aprotic solvent modifiers displayed the most exchange.

This can be explained by envisioning the solvent environment creating a network around the analyte, potentially competing with any deuterated solvent based on gas phase basicity. One idea is that acetone displaces any clustered water or methanol that is interacting with the RNA

competing with VSVI for exchangeable deuterium atoms and decreasing the overall rate of exchange. According to the proposed negative mode HDX mechanism involving nucleosides, it is possible that IPA interacts with the negatively charged phosphate that has been implicated in the mechanism which impedes the overall exchange rate.

1183 1184 1185 1186

Figure 4.6 HDX profiles and difference spectra for the −6 charge state of VSVI in varying modifiers. The −6 charge state was t-infused with D2O and mass spectra were recorded in (top to bottom): isopropyl alcohol, dry nitrogen, acetonitrile, and acetone at CV = 4.6 V and −0.4 V respectively. The difference spectrum is the resultant spectrum when the non-HDX spectrum is subtracted from the HDX spectrum to indicate signal loss or signal gain and number of deuterium atoms exchanged. Raw mass spectra were first fit with a Gaussian smooth before the Origin multi-peak fit analysis was performed and baseline was corrected. Peaks were extrapolated in 0.005 m/z increments from m/z = 1183 to m/z = 1187 before they were normalized by sum to ensure consistency. The fitting procedure and low signal is the reason of the abrupt drop off at high m/z in several of the difference spectra.

The two separated species in the 1.5 % (v/v) acetone modified environment displayed subtle differences in hydrogen-deuterium exchange patterns. The differences are also present in the other charge states, as illustrated in Appendix B. However, it is quite possible that the subtle differences in exchange are due to spectral noise. Based on the postulated differences between active and inactive conformations, it would make sense that the HDX rates are quite similar. A slight increase in deuterium uptake may be related to the protrusion of nucleobases in the active conformation to accommodate substrate loop I, which exposes additional exchangeable

1183 1184 1185 1186 1187

hydrogens. However this may be balanced by newly formed stacking interactions as seen in the C2′-endo conformation involving the protruded nucleobases, such as A₇₃₀⁺ in an interaction with G757 which makes exchange more difficult.

4.4 Conclusions

The above method outlines a potential alternative to traditional NMR and X-ray crystallography techniques to probe the structure and dynamics of large RNA molecules using DMS and quantum chemical calculations. We found that the protonation of a key adenine residue in the active site loop of the Varkud Satellite ribozyme allowed for conformational changes between speculated active and inactive conformations due to electrostatics.

Additionally, sugar repuckering observed on the timescale of NMR experiments was seen in guided computational calculations which allowed for minor groove narrowing and increased stacking interaction between nucleobases to accommodate the substrate loop.

DMS and HDX studies provided fundamental results on how large RNA molecules behave in the gas phase during these experiments in a negative polarity. Two species of VSVI, postulated to be active and inactive conformations, were able to be separated in DMS with the use of an acetone modifier. The rest of the DMS behaviours showed typical clustering behaviours with each modifier, however separation only in acetone was difficult to explain. The use of solvent modifier also had an effect on HDX results, with protic solvents such as IPA impeding the overall exchange rate while aprotic solvents such as acetonitrile or acetone appeared to enhance the exchange rate relative to an unmodified nitrogen environment. We rationalize this by noting that D₂O was t-infused prior to Q1 so solvent modifier, such as IPA, could engage in a hydrogen bond network with the ribozyme and impede the exchange rate while simultaneously competing with phosphate groups for deuterium exchange.

VSVI has a unique dynamic and energetic profile and further study is necessary to fully understand its mechanism. Through the use of DFT calculations, MM simulations, and DMS-MS-HDX experiments, several postulated statements about VSVI function, including nucleobase rearrangement, from NMR experiments were confirmed. Based on the work here, it can be thought that the VS ribozyme is able to catalyze its reaction by key catalytic acidic and basic nucleobases, sugar puckering dynamics, and specific nucleobase protonations to allow for differed electrostatics as highlighted both experimentally and computationally.

5.0 Separation and Identification of Trimethoprim Transformation